Driver License Data are Valuable for Obesity Surveillance

BACKGROUND:  Obesity has emerged as one of public health’s top priorities. Public health agencies need reliable local data for community needs assessments to guide prevention efforts. Existing survey data sources provide county-level estimates; sub-county estimates from survey data are typically too expensive for public health programs to collect. Data from state-issued driver licenses and ID cards (DMV records) can be used for community-level obesity estimates.

METHODS:  We analyzed records of 3.2 million driver licenses and ID cards issued to Oregonians ages 18-84. After geocoding and computing body mass index (BMI) for each individual, we calculated age-adjusted mean BMI for every census block group, tract and county. We compared estimates against the Behavioral Risk Factor Surveillance System (BRFSS) and examined trends for birth cohorts. Address-level data were used for maps with finer resolution than block groups, and to determine proximity to groceries, fast food restaurants and convenience stores.

RESULTS:  Oregon’s DMV database has good data quality, is inexpensive and is easy to analyze. BMI estimates from DMV records averaged 2 percent lower than the BRFSS for men and 5 percent lower for women. Under-reporting of weight in DMV records appears consistent, and the data reveal striking temporal patterns and geographic variation in weight status at the neighborhood level. 

CONCLUSIONS:  DMV records are a valuable resource for obesity surveillance in Oregon. BMI estimates from DMV records will greatly enhance the ability of public health programs and advocates to describe disparities and trends in obesity. Because BMI estimates are consistently conservative, the data are still useful for describing temporal and spatial patterns. Other states should explore using DMV data for obesity surveillance.